Help with gathering insight and inference from profiling data

I was wondering if anyone here has done something where given profiling data for example sweetvz/ydata, you use chatgpt to extract inferences and insight. For example explaining anomalies, or explaining correlations, etc.

I’ve used ChatGPT’s code interpreter to analyse datasets (e.g. .csv files).

You can simply upload the dataset and give a prompt something along these lines?

Here is a dataset about [description]. It contains [description about what each row is - e.g. a response to a survey]. Please [insert analysis - e.g. do some basic EDA to identify correlations between the different columns.

I’ve also asked it to train Machine Learning models (e.g. scikit-learn models which are then downloadable).

1 Like